Evaluation of grid-map sensor fusion mapping algorithms

Abstract

This paper presents a thorough evaluation of grid map based sensor fusion algorithms for mapping the environment of a mobile robot. Three physical sensors were used for creating the grid maps: a CCD camera, a set of ultrasonic sensors and a laser rangefinder. An adaptive fuzzy logic algorithm was compared to three logical sensor fusion algorithms. Results indicate the superiority of the adaptive fuzzy logic algorithm with a 0.075 confidence level.